VOC-alarm: mutation-based prediction of SARS-CoV-2 variants of concern

Hongyu Zhao, Kun Han, Chao Gao, Vithal Madhira, Umit Topaloglu, Yong Lu, Guangxu Jin

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Mutation is the key for a variant of concern (VOC) to overcome selective pressures, but this process is still unclear. Understanding the association of the mutational process with VOCs is an unmet need. Motivation: Here, we developed VOC-alarm, a method to predict VOCs and their caused COVID surges, using mutations of about 5.7 million SARS-CoV-2 complete sequences. We found that VOCs rely on lineage-level entropy value of mutation numbers to compete with other variants, suggestive of the importance of population-level mutations in the virus evolution. Thus, we hypothesized that VOCs are a result of a mutational process across the globe. Results: Analyzing the mutations from January 2020 to December 2021, we simulated the mutational process by estimating the pace of evolution, and thus divided the time period, January 2020-March 2022, into eight stages. We predicted Alpha, Delta, Delta Plus (AY.4.2) and Omicron (B.1.1.529) by their mutational entropy values in the Stages I, III, V and VII with accelerated paces, respectively. In late November 2021, VOC-alarm alerted that Omicron strongly competed with Delta and Delta plus to become a highly transmissible variant. Using simulated data, VOC-alarm also predicted that Omicron could lead to another COVID surge from January 2022 to March 2022.

Original languageEnglish (US)
Pages (from-to)3549-3556
Number of pages8
JournalBioinformatics
Volume38
Issue number14
DOIs
StatePublished - Jul 15 2022

ASJC Scopus subject areas

  • Statistics and Probability
  • Biochemistry
  • Molecular Biology
  • Computer Science Applications
  • Computational Theory and Mathematics
  • Computational Mathematics

Fingerprint

Dive into the research topics of 'VOC-alarm: mutation-based prediction of SARS-CoV-2 variants of concern'. Together they form a unique fingerprint.

Cite this